Sleep is a fundamental aspect of life. When it is routinely disrupted the result can be any number of health conditions or diseases. One of the popular solutions to this is to use a white noise machine. This creates a masking level of background noise to cover sudden disturbances and reduce waking.
The problem I found with these is that consistent use is known to cause hearing damage. As they are commonly used for babies it is an even greater concern. My solution was to develop an algorithm that would intelligently learn from your environment. It then predictively generates white noise at the optimum level for covering sound at that time. This means that at quiet points in the night the output reduces and hearing health is preserved. At noisy times of day, such as at 8am on a Monday morning, the sound increases and sleep is undisturbed.
The system was developed using Raspberry Pi’s with the attached GrovePi sensor board. The noise tracking algorithm was implemented in Python.